see5 0.1.0 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6d2b54a9d4d3c9f3271505bf233a46772cacfd1702312a68dab174b197265f59
-  data.tar.gz: cb5a5d275b08cb8ce7ca00be58c81fd015902578d31c1d6e8a7d1b5d2936225a
+  metadata.gz: f2f7b552c03f5df7d375e4fdf2af7a30bbd25cd4c55634a5be14b5abd0f511f2
+  data.tar.gz: 03056357c74c6cd58145c5a60f9a759b2f8c04fb7809357163900415da1a1791
 SHA512:
-  metadata.gz: 903fb8b64822a99a15c14383fc57932af364bb1c355046b4b022b9588c2d7359c5bdbcf5221f4282ba9b1f9a4200f746d390d962367bcad8a39ecd61e2671839
-  data.tar.gz: 98594a2c2be18ede5633ee7ba66894d2ce2b2eddead9d10eb4531728c349b8b9db95e439f7ca1407d9c92a6abed363549127a7b731a8f5c53af984096d0fe1d5
+  metadata.gz: dd9b563564b17b5b9f2f770448b59ed75947b0ad61fc26377529fed84cca56f5ac31f2b8950635a1a53bd8667d289d19a2cd5cedc2a2783c04b1f485d27cfcfc
+  data.tar.gz: f5b0be6ae50abef89eab752f4008b454a49b00f68bc59bff54a99adcc6968950fc898f43785eca0ee3c4915f75616edbae5c77fd953953a27156976cb51f7206

data/Gemfile.lock

@@ -0,0 +1,22 @@
+PATH
+  remote: .
+  specs:
+    see5 (0.1.0)
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    minitest (5.14.1)
+    rake (13.0.1)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  bundler (~> 1.17)
+  minitest (~> 5.0)
+  rake (~> 13.0)
+  see5!
+
+BUNDLED WITH
+   1.17.2

data/lib/see5.rb

@@ -2,6 +2,38 @@
 
 require "see5/input_file_writer"
 require "see5/model"
-require "see5/rules_file_parser"
+require "see5/rules_output_parser"
+require "see5/gritbot_output_parser"
 require "see5/schema"
 require "see5/version"
+
+module See5
+  def self.train(data, class_attribute)
+    prepare_tmp_files(data, class_attribute)
+    run_see5
+
+    output = See5::RulesOutputParser.parse_file("/tmp/ruby-see5.rules_output")
+
+    See5::Model.new(**output)
+  end
+
+  def self.audit(data, class_attribute)
+    prepare_tmp_files(data, class_attribute)
+    run_gritbot
+
+    See5::GritbotOutputParser.parse_file("/tmp/ruby-see5.gritbot_output")
+  end
+
+  def self.prepare_tmp_files(data, class_attribute)
+    schema = See5::Schema.from_dataset(data, class_attribute)
+    See5::InputFileWriter.write_files(data: data, schema: schema)
+  end
+
+  def self.run_see5
+    system("c5.0 -f /tmp/ruby-see5 -r > /tmp/ruby-see5.rules_output")
+  end
+
+  def self.run_gritbot
+    system("gritbot -s -f /tmp/ruby-see5 -r > /tmp/ruby-see5.gritbot_output")
+  end
+end

data/lib/see5/gritbot_output_parser.rb

@@ -0,0 +1,92 @@
+# frozen_string_literal: true
+
+module See5
+  # Read Gritbot output and return an array of hashes representing the anomalies
+  class GritbotOutputParser
+    attr_reader :anomalies
+
+    def self.parse_file(fname)
+      new(fname).anomalies
+    end
+
+    def initialize(fname)
+      @file = File.open(fname)
+      @anomalies = []
+
+      parse_file
+    end
+
+    def parse_file
+      discard_header
+
+      while (line = lines.next)
+        if line.start_with?(/\s*while checking/)
+          # TODO record excluded cases
+        elsif line.start_with?(/(:?test |data )?case /)
+          @anomalies << parse_anomaly(line)
+        elsif line.start_with?("Time:")
+          break
+        end
+      end
+    end
+
+    private
+
+    def lines
+      # TODO: lazy unnecessary given that rules are small?
+      @file.each_line.lazy
+    end
+
+    # Discard the file header and advance to the anomalies section
+    def discard_header
+      while (line = lines.next)
+        break if line.start_with?(/\d+ possible anomal/)
+      end
+      # discard the final blank line
+      lines.next
+    end
+
+    def parse_anomaly(line)
+      info = parse_anomaly_info_line(line)
+      value = parse_anomaly_value_line(lines.next)
+
+      conditions = []
+      while (line = lines.next.strip)
+        break if line == ""
+
+        conditions << parse_condition_line(line)
+      end
+
+      # TODO new class for these
+      {
+        **info,
+        **value,
+        conditions: conditions
+      }
+    end
+
+    def parse_anomaly_info_line(line)
+      matches = line.match(/.*case (.*): (.*)\[([\.\d]+)\]/)
+
+      {
+        case_index: matches[1],
+        case_label: matches[2].strip.delete_prefix("(label ").delete_suffix(")"),
+        signifigance: matches[3].to_f
+      }
+    end
+
+    def parse_anomaly_value_line(line)
+      matches = line.match(/(.*) = (\S*)\s*\((.*)\)/)
+
+      {
+        attribute: matches[1].strip,
+        value: matches[2],
+        reason: matches[3]
+      }
+    end
+
+    def parse_condition_line(line)
+      line.strip
+    end
+  end
+end

data/lib/see5/input_file_writer.rb

@@ -20,6 +20,9 @@ module See5
     def write_files
       write_names_file
       write_data_file
+
+      names_io.close
+      data_io.close
     end
 
     def write_names_file
@@ -38,7 +41,12 @@ module See5
 
     def row(record)
       @schema.attributes.map do |attr, _vals|
-        record.send(attr)
+        if record.is_a?(Hash)
+          record[attr]
+        else
+          # assume some kind of OpenStruct- or ActiveModel-like object
+          record.send(attr)
+        end
       end.join(",")
     end
 

data/lib/see5/{rules_file_parser.rb → rules_output_parser.rb}

@@ -6,7 +6,7 @@ module See5
   # Read See5 rules output and return an array of hashes representing the rules
   # Note that this is the output normally sent to stdout, NOT the .rules file!
   # The .rules file lacks some important information like confidence.
-  class RulesFileParser
+  class RulesOutputParser
     def self.parse_file(fname)
       new(fname).model
     end

data/lib/see5/schema.rb

@@ -22,5 +22,25 @@ module See5
           "#{attr}: #{vals.join(',')}"
         end.join("\n")
     end
+
+    # Infers a schema from a dataset
+    def self.from_dataset(data, class_attribute: nil)
+      classes = data.map { |record| record[class_attribute.to_sym] }.uniq
+
+      attributes = {}
+      data.each do |record|
+        record.each do |key, value|
+          if !attributes.include?(key)
+            attributes[key] = [value]
+          elsif !attributes[key].include?(value)
+            attributes[key].append(value)
+          end
+        end
+      end
+
+      new(classes: classes,
+          attributes: attributes,
+          class_attribute: class_attribute)
+    end
   end
 end

data/lib/see5/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module See5
-  VERSION = "0.1.0"
+  VERSION = "0.2.0"
 end

data/see5.gemspec

@@ -34,5 +34,5 @@ Gem::Specification.new do |spec|
 
   spec.add_development_dependency "bundler", "~> 1.17"
   spec.add_development_dependency "minitest", "~> 5.0"
-  spec.add_development_dependency "rake", "~> 10.0"
+  spec.add_development_dependency "rake", "~> 13.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: see5
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
 platform: ruby
 authors:
 - Eddie Lebow
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2020-03-03 00:00:00.000000000 Z
+date: 2021-01-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -44,14 +44,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '13.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '13.0'
 description: 
 email:
 - elebow@users.noreply.github.com
@@ -62,13 +62,15 @@ files:
 - ".gitignore"
 - ".rubocop.yml"
 - Gemfile
+- Gemfile.lock
 - README.md
 - Rakefile
 - lib/see5.rb
+- lib/see5/gritbot_output_parser.rb
 - lib/see5/input_file_writer.rb
 - lib/see5/model.rb
 - lib/see5/rule.rb
-- lib/see5/rules_file_parser.rb
+- lib/see5/rules_output_parser.rb
 - lib/see5/schema.rb
 - lib/see5/version.rb
 - see5.gemspec